Accurately measuring the cup angle following THA and comparing it with the "safe zone"2,3 is crucial for assessing surgical outcomes and guiding surgeons, particularly younger ones, in enhancing their surgical techniques. Moreover, for patients experiencing recurrent dislocation post-THA, precise measurement of the cup angle can inform treatment decisions. Various methods for measuring anteversion have been documented in the literature. The Lewinnek’s method2, for instance, employs trigonometric functions to calculate the major and minor axes of the projected ellipse to determine anteversion. Pradhan's method4 utilizes distinct reference points and lines within the ellipse. Liaw 6 developed a protractor for use on radiographs, simplifying the process of measurement and calculation. Widmer5 observed that within a specific range, the ratio of the ellipse’s major to minor axis is nearly linearly related to the anteversion angle, leading to the design of a simpler method based on Lewinnek's technique. Callanan 3 utilized the specialized Martell Hip Analysis Suite™ software (Chicago, IL) to analyze hip radiographs and establish the well-known "Callanan safe zone" for acetabular angles.
The methods described above measure the RA, typically through the analysis of parameters associated with the ellipse projection on AP hip radiographs, thereby estimating the RA indirectly. During practical application, however, we identified several limitations with these methods:
1. Positional Variability: While some studies have validated the high accuracy of these methods 5, 8, 12, most validations were conducted using in vitro models. Clinically, patient positioning can deviate from the ideal due to factors like pain, making it challenging to maintain the precise alignment required for accurate measurements. Patients may rotate their bodies instead of aligning perfectly with the X-ray beam, leading to measurement discrepancies. 2. Ellipse identification: Identifying the projected ellipse is a critical initial step. However, in practice, the complex structures of the prosthetic components, such as the metallic edges, coatings, and grooves designed to secure the liner, can create multiple overlapping ellipses on AP radiographs. These projections, further obscured by the femoral head prosthesis, bone, and soft tissues, appear as a series of confusing arcs, making it difficult to precisely define the ellipse and establish the necessary reference points and lines. 3. Trigonometric transformation amplifies errors: Due to the reliance on trigonometric transformations, even minor inaccuracies during measurement can amplify into great errors in the final RA calculations. 4. Anteversion vs. retroversion determination: AP radiographs alone cannot distinguish between anteversion and retroversion, necessitating additional imaging modalities such as CT scans or lateral radiographs. While some of these limitations have been addressed by advanced software solutions, their use is not widespread, and the improvements offered are limited.
Yao 13 measured the RA on a cross-table lateral radiograph. In this procedure, the X-ray is directed perpendicularly to the OBE plane (as depicted in Fig. 1), resulting in a radiograph parallel to this plane. The RA is defined as the angle between the cup axis and the long axis of the body in the cross-table lateral view. However, cross-table lateral radiographs are not commonly used in clinical practice due to their complex execution, which involves multiple radiations and adjustments of the X-ray direction. Ghelman 10 noted that this method has low accuracy and yields highly variable results even when the same observer measures different cross-table lateral radiographs of the same patient.
Numerous studies have quantified CT measurements (as depicted in Fig. 2b). For instance, a retrospective analysis of 42 patients by Ghelman 10 revealed higher intraobserver and interobserver reliability for CT compared to AP radiographs, as well as independence from variations in patient positioning. It is important to note, however, that the anteversion angle measured via CT is specifically referred to as the AA. Unfortunately, there is currently no established safe zone for AA to assess acetabular angles. This distinction is often overlooked in both clinical practice and literature, where AA and RA are frequently conflated. As illustrated in Table 3, the relationship between AA, RA, and RI (as defined by Eq. 1) demonstrates that AA and RA are not equivalent, with significant differences between them. Thus, it is clearly inappropriate to compare the AA obtained from CT scans with the RA safe zones proposed by Lewinnek2 and Callanan3. In reviewing the association between acetabular cup position and dislocation, Seagrave7 observed that various studies employed differing definitions and methods for measuring anteversion. This inconsistency represents a significant confounding factor when comparing acetabular anteversion against target values. Seagrave recommends adopting a more standardized approach to measurement.
Table 3
Examples of the mathematical relationships between the RI, RA and AA.
| RI(°) | 30 | 30 | 30 | 40 | 40 | 40 | 50 | 50 | 50 |
| RA(°) | 5 | 15 | 25 | 5 | 15 | 25 | 5 | 15 | 25 |
| AA(°) | 9.92 | 28.19 | 43.00 | 7.75 | 22.63 | 35.96 | 6.52 | 19.28 | 31.33 |
In studies employing in vitro models5, 8, 12, researchers can directly ascertain the anteversion angle as a standard value. However, in both this study and clinical practice, the actual cup anteversion angle within a patient's body cannot be directly obtained. The 3D method in this study was implemented by precisely defining the RA in a digital model that accurately mirrors the pelvic structure. Consequently, we posit that the measured values from this method can be considered a reliable reference for the actual RA. The extremely high intraobserver and interobserver ICC of RA3Ds further support the accuracy and reliability of the 3D method. The 3D method notably remains unaffected by the patient's position. However, it does have its limitations: it requires specialized image processing software, which means observers must undergo training; the procedure can be intricate, and the use of 3D imaging contributes to higher costs.
The new method offers a more straightforward operation and is less influenced by the position of the pelvis compared to existing techniques. AP radiographs and CT scans are standard clinical procedures that do not add additional costs. Our study demonstrated that the new method boasts higher intraobserver and interobserver reliability as well as greater accuracy than the traditional approach. It is important to note, however, that the new method is not entirely immune to variations in pelvic position. While the measurement of the AA on CT scans remains unaffected by pelvic rotation, the measurement of the RI on AP radiographs does show some sensitivity. Using mathematical models, we find that for a patient with an actual RI of 40° and a real RA of 20°, a 5° rotation of the pelvis toward the unaffected side (resulting in the affected hip moving forward and the healthy hip moving backward) would yield a theoretically measured RI of 38.47° on AP radiographs. The theoretical RA calculated using the new method would be 19.41°, a deviation of only 0.59° from the actual value. In contrast, the theoretical RA measured using Lewinnek’s and Pradhan’s methods, without considering other potential measurement errors, would be 23.16°, representing an increase of 3.16°. This clearly indicates that changes in pelvic rotation have a minimal impact on the new method.
We observed that recent years have seen a growing body of literature examining the impact of pelvic motion and tilt on acetabular orientation, while challenging the traditional concept of the "safe zone"14,15. The new method proves equally effective for measuring RA in both standing and sitting positions, provided that the AP and CT data are complete.
Shortcomings: While this study improves the accuracy of measuring RA utilizing basic clinical data such as AP radiographs and CT scans, when compared to traditional methods, it is worth noting that in scenarios where advanced data, software, and tools are available, including 3D imaging data and even joint surgical robotic systems16, other methods may potentially achieve even higher levels of accuracy.